Southwest Indian Ridge (Fig. F1) extends 7200 km from the Bouvet Triple Junction to the Rodriguez Triple Junction and has been in existence since the breakup of Africa and Antarctica (Norton and Sclater, 1979). Half-spreading rates over the last 34 m.y. calculated for the Southwest Indian Ridge at the Bouvet Triple Junction (Lawver and Dick, 1983) and near the Indian Ocean Triple Junction (Tapscott et al., 1980) are 0.78 and 0.8 cm/yr, respectively. Thus, the SWIR is at the very slow end of the spreading rate spectrum (Fisher and Sclater, 1983). Various features of slow-spreading ridges including deep rift valleys, rough topography, and abundant exposures of plutonic and mantle rocks are all typical characteristics of Southwest Indian Ridge (Fig. F1).
The Atlantis II Fracture Zone is one of many large-offset and high-relief fracture zones offsetting the Southwest Indian Ridge axis (Dick et al., 1991; Meyer et al., 1989). It has a left lateral offset of 199 km, which represents a ~20 m.y. offset of the SWIR (Engel and Fisher, 1975). The transform extends approximately north-south along longitude ~57°E (Fig. F1). It has a very deep (>6 km) transform valley (Dick et al., 1991) and contains a median tectonic ridge and high flanking transverse ridges on both sides of the transform (Fig. F2). The valley walls are extremely steep for large distances and are covered with extensive debris and talus. The ridge transform intersections (RTIs) are marked by deep nodal basins, lying on the transform side of the neovolcanic zones that define the present-day spreading axes, and intersect clearly defined rift valleys with a relief of >2200 m and widths of 22–38 km (Dick, Natland, Miller, et al., 1999).
Hole 1105A is located on the east side of the Atlantis II Transform Fault, which offsets the Southwest Indian Ridge. It is situated at a unique outcrop of lower oceanic crust exposed on a 15-km-long wave-cut terrace (Fig. F3) at 730 m water depth (Dick et al., 1991). The platform exposures of plutonic rock are thought to have formed at ~11.5 Ma as a core complex at the northern RTI as part of the Antarctic plate (Dick et al., 1991). Schwartz et al. (2005), utilizing Pb/U ages, determined that the Atlantis Bank likely formed between 11.2 and 12.5 Ma, very similar to the proposed magnetic ages determined. Hosford et al. (2003) showed through analysis of magnetic data from ridge segment AN-1 (north of the Atlantis Bank) out to 25-Ma lithosphere (south of Atlantis Bank) that spreading rates during formation of the Atlantis Bank were highly asymmetric. Half-rates of ~0.85 cm/yr were documented to the south and 0.55 cm/yr to the north of ridge segment AW-1. At the time of formation the gabbroic and ultramafic rocks outcropping on the Atlantis Bank are thought to have been unroofed and exposed on a low-angle detachment fault on the southern flank of the rift valley near the RTI that was subsequently partially uplifted above sea level and then transported to its present position approximately halfway along the transform. Analysis of the seafloor fabric around the Atlantis Bank shows that the platform margins display a north-south seafloor fabric truncated north and south of the bank by an east-west spreading tectonic fabric (Fig. F4). We interpret the Atlantis Bank fabric to possibly represent remnants of a corrugation fabric typical of core complexes or megamullions observed in the Atlantic Ocean (e.g., Tucholke et al., 1997), although this is less clear because part of the massif was above sea level and eroded (Dick et al., 1991). Figure F4 is a three-dimensional (3-D) view of the transform domain with a false-color illustration of the north-south transform-parallel seafloor fabric. The fabric is defined by first creating a slope map from an unfiltered bathymetric grid on the Atlantis Transform and then defining the aspect of slope. The aspect map is draped on bathymetry, and the image depicts color changes at positions of maximum slope or changes in slope direction. The trends of the color changes are parallel to the steepest slopes and axes of slope direction change. Linear morphology of large- or small-scale structures such as corrugations or faults and their continuity become readily apparent, as highlighted by color changes. The transform-parallel fabric appears to be very distinct from elsewhere along the Atlantis Bank transverse ridge where the fabric typically parallels the spreading center. The transform-parallel corrugations manifested by the aspect map appear to persist on the lower part of the eastern wall of transform valley.
Baines et al. (2003) inferred that the same north-south fabric may be caused by more recent transform-parallel normal faulting of unknown origin, in part to explain the excess elevation of the transverse ridge and in particular the Atlantis Bank. We are, however, struck by the fact that these transform-parallel seafloor fabric elements over the eastern transverse ridge are truncated by east-west spreading center–parallel faults and seafloor fabrics just north and south of the platform, as is typical over many of the core complex counterparts that have been studied in the Atlantic (Tucholke et al., 1998). If large ridge-parallel faults originate nearer the ridge axis, as seems likely, and the transform-parallel corrugations are cut by the ridge-parallel faults, it is likely the corrugations originated near the ridge axis and not off axis, as inferred by the model of Baines et al. (2003). That the steep parts of the mullion corrugations could nucleate high-angle transform-parallel normal faults subsequent to unroofing is certainly possible. Gravitational collapse along a normal-slip surface parallel to the transform wall is also possible, as are normal slip surfaces related to differential subsidence rates on seafloor adjacent to the transform, as predicted by Fox and Gallo (1984). The stress field necessary to create nearly transform-parallel pure normal slip faults as suggested by Baines et al. (2003), other than collapse caused by high relief, is indeed somewhat uncertain. The excess elevation of the transverse ridge and local normal slip faulting, in part, could be aided by bending strains and extensive subsurface volume expansion of 20%–40% caused by serpentinization of mantle material. Materials collected during submersible dives and dredges and drill core taken along the transform valley walls adjacent to Atlantis Bank have recovered suites of highly serpentinized ultramfic rocks (MacLeod et al., 1998). Sustained tectonic activity along the transform could lead to fracture-assisted fluid penetration in the regions adjacent to the transform and enhanced hydration and transformation of mantle material to crustal material and densities. This may, in part, explain some of the excess elevation along the transverse ridge.
It is important to note that our FMS analyses of fractures and shear zones in Hole 1105A and those of Hole 735B (Iturrino et al., 2002) show that these fractures and shear zones typically strike parallel to the northern ridge axis approximately east-west and have normal senses of slip. These shear zones and faults appear to have a conjugate relationship with moderate dips either toward the ridge axis (north) or away from the ridge axis (south), which may be due to normal inward-dipping ridge axis faults as well as conjugate faults resulting from bending stresses during footwall rollover (e.g., Tucholke et al., 1998). Transform-parallel normal fault trends inferred by Baines et al. (2003) are represented in the core. Likewise, Iturrino et al. (2002) show that fault plane solutions from recorded earthquakes in the region are either north-south-striking strike-slip faults or east-west-striking normal faults, with an absence of north-south-striking normal fault solutions. The 3-D image shown by Baines et al. (2003) could alternatively be inferred to show a typical transform-parallel mullion corrugation structure on the uneroded flanks of Atlantis Bank.
Formation of low-angle detachment faults and core complexes such as the Atlantic Bank are generally thought to be the consequence of amagmatic or magma-poor extension at the ridge and development of a master fault, which lead to exposure of a plutonic or serpentinized mantle section of the oceanic lithosphere (Fig. F4). The core complex effectively forms a tectonic window (Karson et al., 1987; Karson and Winters, 1992; Cannat and Casey, 1995; Tucholke et al., 1998) into the lower oceanic crust or upper mantle exposed on the footwall of a now low-angle detachment. Holes 1105A and 735B appear to be drilled on such a tectonic window on the Atlantis Bank (Dick, Natland, Miller, et al., 1999). Hole 1105A was located to help constrain the overall structure of the massif exposed on the platform and to compare with the upper regions of Hole 735B. Drilling in Hole 1105A provides an opportunity to understand the dimensions of subaxial magmatic systems and continuity of structure, lithology, and magmatic facies. It also provides the opportunity to constrain magmatic and structural processes along strike of the ridge axis at a very slow spreading center. These opportunities are based on possible similarities and correlations between Hole 1105A and the upper part of Hole 735B, holes with ~1.2 km of lateral offset (Fig. F3).